A simple visualization tool is provided. It comes packaged as an executable jar, or you can download the source. As in the past, to use this tool you must create an executable program out of your TopCoder submission. This executable should read the image from standard in and write your segmentation to standard out. You should first read an integer S, on a single line by itself. The next S lines will then contain the S rows, as they would be provided by the TopCoder interface. After reading this input, you should write exactly S lines, each with S characters, representing the S elements of your output. For example, you could implement the following wrapper for your recognize method.
    main()
        S = int(nextLine())
        for(i = 0; i<S; i++)
            input[i] = nextLine()  
        output = recognize(input)
        for(i = 0; i<S; i++)
            writeLine(output[i])
To run your program, you should execute:
java -jar Textures.jar -seed <seed> -exec <exec>, where <seed> is the random number generator seed used create the test case (the seeds in the examples are 1-10) and <exec> is the executable command to run your program. For some languages this will simply be an executable file. For others, such as Java, this is the command to execute your compiled code. If <exec> contains spaces it must be surrounded by quotes. Thus, if you compile your Java code into a class Main, you would run the second example by doing:
java -jar Textures.jar -seed 2 -exec "java Main"

You may optionally add a '-novis' flag to run the test without displaying the result. Furthermore, if you omit the -exec parameter, the test case will be displayed but not run.

To assist in your debugging, anything you write to standard error will be forwarded to standard error. Because you are writing your segmentation to standard out, it is important that you not write anything else.

As a final note, be sure you flush your buffers after writing any output, to ensure that the visualizer receives the data.